Background: Ensuring timely access to mental health services remains a major challenge for VA. In particular, access to psychiatrists was identified as a critical bottleneck by the Office of Inspector General (OIG)'s 2012 Review of Veterans' Access to Mental Health Care. The OIG's 2015 follow-up audit found that many VA facilities do not have explicit psychiatrist staffing plans. It is therefore important for facilities to (i) optimize the allocation of available outpatient psychiatrist appointment slots across Veterans such that those in need of more intense care can be seen more often by their psychiatrists compared to others who are in less need of active treatment or frequent monitoring, and (ii) accurately plan psychiatrist staffing levels to match the optimal appointment frequency needs of the population of Veterans that they treat. Significance/Impact: Frequency of mental health appointments is predefined for VA-endorsed evidence- based psychotherapies (e.g., once-a-week for 12 weeks). However, optimal psychiatrist appointment frequencies for Veterans who are not, or no longer, undergoing such structured therapies are not well established. We recognize that the appropriate appointment frequency for a Veteran ought to be a case-by- case clinical decision based primarily on the psychiatrist's expertise and the Veteran's preferences, supported by best available evidence. However, there exists little evidence on which appointment frequencies can be safely and appropriately considered for which Veterans. This pilot launches a program of research for supporting optimal allocation of care resources, to effectively enhance access and improve outcomes. This proposed pilot research thus (i) directly addresses VA's priorities of Improved Timeliness and Efficiency as outlined in its Fiscal Year 2018-2019 Operational Plan, and also (ii) facilitates the use of VA data as a national resource, which is currently set forth as one of VA Research's top priorities. Innovation: Our work will specify mathematical optimization functions/parameters to construct computational models that provide information to innovatively support mental health care planning. Based on findings from this pilot, a subsequent IIR will develop, implement, and evaluate novel decision support tools. Specific Aims: Aim 1: Systematically identify an expert panel-endorsed (i) definition of stable Veterans to include in this investigation (e.g., treated in the general mental health clinic and psychiatrically stable with no psychotropic medication changes over six months) and (ii) definition of Veteran-level mental health outcomes (e.g., psychiatric emergency room visits / hospitalizations, major medication changes, medication adherence) to utilize in Aim 2 for risk assessments based on CDW data. Aim 2: Computationally analyze optimal psychiatrist appointment frequencies to develop a risk-stratified list of minimum psychiatrist appointment frequencies that are unlikely to result in a negative change in mental health outcomes for stable Veterans, based on the Veteran risk profiles (e.g., different clinical profiles) and outcomes defined under Aim 1, while accounting for site-level differences and non-psychiatric care utilization. Methodology: Guided by safety engineering's risk-based decision-making model, we will utilize a moderated expert panel process (Aim 1) and mathematical optimization (Aim 2) for our aims, targeting psychiatrically stable Veterans who are potentially in less need of active mental health treatment or frequent monitoring. Next Steps/Implementation: A subsequent IIR will introduce and examine the use of this evidence by the field ? namely, whether or not our evidence-based risk-stratified list of appointment frequencies is able to support (not mandate) decision making on the part of (i) VA psychiatrists and stable Veterans in making appointment frequency decisions and (ii) VA mental health services in making psychiatrist staffing plans. We will also concurrently prepare for implementation by identifying barriers to and enablers of implementing these tools and by engaging stakeholders and operational partners in implementation planning.